The Journal. - Is an Algorithm Blocking Your Job Search? One Man Is Trying to Find Out.

Episode Date: August 14, 2025

Derek Mobley spent years applying for jobs online after he got laid off. After more than 100 rejections, he started to wonder whether Workday – the popular online recruiting platform – was to blame. Now, Mobley is suing Workday, alleging that the company’s hiring algorithm discriminated against him. Derek shares his experience, and WSJ’s Lauren Weber explains how his lawsuit got this far. Annie Minoff hosts. Further Listening: AI Is Coming for Entry-Level Jobs What's the Worst AI Can Do? This Team Is Finding Out. Sign up for WSJ’s free What’s News newsletter.  Learn more about your ad choices. Visit megaphone.fm/adchoices

A few years ago, Derek Mobley was working in IT. He loved his job so much. He told me he happily commuted an hour and a half each way to his office in Atlanta. But about a year into the job, Derek got called into a meeting. Next thing I know, you know, they say, hey, it was time to lay some people off. Nothing personal. It's just your numbers come up. And it was just time for me to go.

And that was probably the most difficult day that I could recall like in my adult life because it was a job I really did love. Derek was laid off in 2017. After that, he picked himself up and embarked on a common and frustrating experience, the modern job hunt. And how many applications would you say that you submitted? You know, I know it was probably well over 150 opportunities that I've applied for. How many callbacks do you?

did you get? Of those, really none. Wow. Yeah, yeah. How did that feel? Terrible. I mean, you know, financial stress, emotional stress, I just kept getting a lot of notes. It was just rejection after rejection after rejection.

It was demoralizing. But Derek kept at it, sending out application after application for nine months. At that point, he noticed something. One day, I was looking at my phone. and I got an email, and it was at a very odd time of the day. It was like 1.30 in the morning, and it was like on a weekend, and it was like, okay, Sunday went there. Like, who's actually looking at my resume on my application at that time of day and that time on the weekend? That basically was the watershed moment.

Getting that email at a weird time of the morning, and I started thinking, hmm, this is not a human. This is a bot. Yeah, this is definitely a bot. thought. Derek realized that an algorithm could be playing a big role in all those rejections he was getting. And that realization raised a lot of questions for him. Just what were those algorithms doing exactly? And why couldn't he seem to catch a break? Those questions would eventually take Derek all the way to court, in a lawsuit that could potentially impact millions of job seekers. Welcome to The Journal, our show about money, business, and power.

I'm Annie Minoff. It's Thursday, August 14th. Coming up on the show, one man's journey to find out if algorithms got in the way of his job search. Every year, Americans looking for work fill out millions of job applications. And our colleague Lauren Weber says that nearly all of them get submitted online. Even if you walk into a store these days looking for an hourly job, I recently went through this with my 17-year-old son, you can't fill out a paper application anymore. you get told to go to the website and fill out the application online. These days, being online often means having to interact with artificial intelligence in one way or another. Lauren says the job hunt is no exception.

Lauren, how common is it for employers to use some kind of AI in recruiting? It is ubiquitous, and we use artificial intelligence now very broadly. But in this situation, we're not talking about gender. generative AI like chat GPT, we're talking about algorithms. The basic system is a simple keyword match. Say you're applying to a job in digital marketing. The software will look for the words digital and marketing and for synonyms of those words. And they're going to score and sort applications based on how well a person's application

fits with the words that are in the job description. One of the biggest platforms using this kind of technology is Workday. Workday offers tools for managing HR, payroll, and recruiting. Tens of thousands of employers around the world, including the Wall Street Journal, use Workday, which means that the platform is often the first point of contact for job seekers filling out an application. Derek Mobley was one of those people. It was just kind of interesting. It was like, wow, it's a lot of companies that utilize this platform.

When Derek was on the job hunt after getting laid off in 2017, he kept finding himself on workday. Normally, what I see is some type of workday moniker, either it's in the URL. It might say ABC Company.orgday.com or it just might have the company name, but then it'll have a workday branding on the actual landing page that I'm on to try to apply for the position. And what kinds of things do you have to submit? Yeah, usually resume and voluntary questions that they ask you, demographic-wise, race, ethnicity, disability. Lauren says that companies ask for this demographic information for their own equal employment record keeping, which some employers have to report to the government. And sometimes, Derek says employers asked other questions.

They're asking you, like, maybe your attitude about certain things, you know, dealing with customers or how you interact with people, how you respond to the coworkers. just things that nature. And I guess obviously it's tailored to the company that you're applying for. Workday says that applications on its platform are unique to each employer, and that it doesn't own or administer tests to evaluate a person's intellect, capabilities, or personality. Once a job seeker like Derek submits an application through Workday, the company says its algorithm slots them into one of four categories,

a strong, good, fair, or low match for the position. Workday says we put them into these categories of the buckets based on how they score, and then we give that information to the actual employer, Workday's customers. And from there, Workday would say we believe in ethical technology, ethical artificial intelligence. We believe there should be a human in the loop is the phrase that's often used. But for candidates like Derek, those hiring decisions can feel awfully automated and impersonal. These emails, they never come with any signatures. It's always, you know, the human resources department.

It's never a specific person. Back when I was younger, there was usually a name associated with it, you know, like, you know, Bob from human resources, you know. Bob didn't think you were a fit. Right. You know, and the most frustrating thing was there were never any initial, you know, callbacks. There was never pre-screening. And usually pre-screening is very cursory, you know, is you have these certain criteria that kind of met certain criteria for opposition. hey, let's have a conversation.

And that wasn't happening. Over a two-year period of unemployment, Derek spent hours applying for more than 100 jobs through Workday. Jobs he felt qualified for, if not overqualified. Still, the rejections kept coming. And that made Derek wonder if there was something more going on, something to do with Workday's algorithms. The more that I kept digging, the more emails I pulled up,

just a frustration of that, I'm thinking, There has to be, there's something here. There's, there's, there's, there's, I'm smelling the smoke. Not necessarily seeing the fire, but I'm definitely smelling the smoke. What do you think these algorithms might be picking up on? I mean, I think it's a combination of things. There are so many different things that people can glean from like your resume and things that you put in the application, disability, age, you know, race.

For example, Derek says he suffers from anxiety and depression, and he sometimes acknowledged that in his own. applications. What do they ask you in those voluntary forms? If you have a disability or not, there were times that I did indicate I had a disability. Again, employers often gathered this information for their own equal employment record keeping. Derek also wondered if Workday's algorithms were picking up on his race.

You know, it could be a race thing. I'm not 100% sure. I went to a historically black university, or college, I should say. And he also thought his age could be because. coming into play. Because I noticed as I got older, the less opportunities I received for interviews, the less responses I received. Then it got to a point where I started to take the date of graduation off my resume. See, maybe that might help. You were trying to game the algorithm through what you submitted. Yeah, I was trying to do the best I can to get a fair shake,

you know, not trying to beat a system, not trying to do anything improper, just to get a fair shake. to get somebody to get a chance to look at what I bring to the table. Workday says that its recruiting tools aren't trained to use or even identify protected demographic information like race, age, or disability. The company also noted that its customers configure its systems to fit their own needs, and that Workday doesn't control how or when employers might update job candidates about their applications. Still, Derek had become convinced he might have a legal case. What I'm saying, you know, there is some kind of inherent bias.

I find it mighty strange that no matter what job it was, no matter what company it was, no matter what the position was, it was always a automatic decline when I went through the Workday platform. And so, Derek sued Workday. So the case was filed in 2023. That's Lauren again. And I noticed it probably right. around the time when it was filed, was kind of interested. I, like many people that I've since spoken to about it,

sort of assumed it wouldn't get very far. But it did. Where Derek's case stands now is after the break. Derek Mobley sued Workday in 2020. His lawsuit claims that the company's hiring algorithm discriminated against him based on age, race, and disability. It is somewhat a novel type of case because usually people sue the employer

that they didn't get the opportunity from. They usually don't sue workday. And my contention is that it's the platform that's part of the problem. It's not necessarily saying that workday in and of itself a discriminatory, it's they're providing the tool for discrimination. You know, it's kind of like people providing the weapons that cause the destruction. In a statement, Workday said that Derek's lawsuit is, quote, without merit, adding that its recruiting tools don't make hiring decisions, which are always made by its customers.

Workday also disputed that it provides tools for discrimination. So they are trying to say, if Derek Mowbly is not getting jobs, it's not because of our software. Workday has tried to get Derek's case thrown out. And in an early ruling, the judge in the case agreed that Workday's algorithms were not intentionally discriminatory. But she also left a door open for Derek and his legal team. Now, even if Workday didn't intend to discriminate, he can still try to show that Workday discriminated in the impact, in the effects of its algorithms. That's called disparate impact. So he doesn't have to prove, hey, your algorithm is discriminating against me.

He just has to prove it has had the effect. of discriminating. It's the outcome that matters. It's the outcome that matters. In May, without addressing Derek's race and disability claims, the judge said that Derek's case could proceed on the basis of age discrimination.

Experts Lauren talked to told her that proving workday's algorithms had the effect of discriminating would most likely involve looking at the platform's internal data. For example, when companies have any kind of product or a test, certainly a hiring test or software, they should be doing some analysis to make sure that it is not having this disparate impact, that it's not affecting people of certain

groups in a negative way. They may have already done that. And so this may just be a question of them producing the tests that they've already done. There's a lot of discovery issues that they're going to be wrangling over, meaning how much information does Workday have to show So Derek and his lawyers, any software company will argue that its algorithms are proprietary. They don't want any of this to become public. They can't share it. These are trade secrets. How much do they have to divulge?

There's a lot of complicated legal negotiating ahead. We asked Workday whether it planned to share data with Derek and his team. And the company said it would be inappropriate to comment, given that the matter will soon be discussed in court. Derek is hopeful. I don't need to personally know how the algorithm works. I don't need to see the code or any of that stuff. But I think that people who are trusted experts should have the opportunity to do so. And people who are independent, you know, and they can verify, oh, there's nothing to see here.

Everything is fine. You know, it's probably just his bad luck. Oh, then I can accept that and move on in my life. After a lengthy search, Derek did finally get a job. In 2019, he moved into a new field, insurance. How did you eventually land your current job? Well, I mean, after going through a lot of stuff, divorced and relocating and things that nature, I ended up in Charlotte, and one day I got a call out of the blue.

It was from a recruiter. A person. A live, bona fide person. She was from Allstate, and she mentioned that there was an opportunity that they were looking to fill, and she set up for the initial pre-screen over the phone, did that after the initial pre-screen, now it's supposed to come into the office, interview with them in person, and end up getting a position.

So this really came through kind of the old-fashioned way. Exactly. You talked to a person, you went into the office. Exactly. And boom. Since then, Derek's been promoted twice. He likes his job, but says he's still making up for the time he's been unemployed and living off his retirement savings.

His case against Workday may go through years of legal wrangling. Right now, the case has cleared the first step to becoming a class action. At the point that it becomes a class action, if it gets that far, anybody who has applied for a job who's over the age of 40, who was essentially denied a job, and they had applied through Workday's platform, potentially has a claim against Workday. As Workday pointed out in its arguments about this,

that could be many millions of work. People. Lots of people. Lots and lots of people. Any job seeker over the age of 40, and I believe it's a five-year period from like 2020 to 2025, you know, certain dates within that. And that was significant because if this, you know, continues to move forward, Workday could have millions of claims on its hands. Workday continues to believe that the case is without merit and says that it shouldn't be certified as a class action. Regardless of the eventual outcome of the case, it seems to have already struck a nerve with some people who feel burned by how hiring works these days. After Lauren's story published, she heard from lots of readers who worried about the role that algorithms play in hiring. I think people are looking for some assurance, some assurance that the process is fair. And after that story came out, I was kind of bowled over by the, The number of people I heard from who said, I'm convinced that these systems discriminate against me because of my age. There are a lot of people out there who are convinced that the job market is not fair.

And I think that's why Derek became sort of a hero to many of our readers who saw that story and wrote to me to say, not only are they pulling for Derek, but how can I join his lawsuit? That's all for today, Thursday, August 14th. The Journal is a co-production of Spotify and the Wall Street Journal. If you like our show, follow us on Spotify or wherever you get your podcasts. We're out every weekday afternoon. Thanks for listening. See you tomorrow.